Do Dietary Intake and Blood Level of Total Selenium Predict Circulating Levels of Selenium Species? A Cross-Sectional Study

Abstract

Some of the difficulties to assess exposure to the metalloid selenium in epidemiologic studies arise from a typical feature of this metalloid, its presence in both the environment and the human body in different chemical species, which in turn have markedly different toxicological and nutritional properties. Very little is known, in particular, about the ability of overall dietary selenium intake and blood selenium level to predict exposure to the various species of this element in the human. We investigated this issue by assessing through a food frequency questionnaire the total selenium dietary intake and by determining plasma selenium level in fifty adults randomly drawn from the municipal population of Modena, northern Italy, and by comparing their values to the plasma concentrations of the various selenium compounds obtained through speciation analysis. Dietary intake did not correlate with total plasma selenium neither to the various selenium species, except for a weak direct association with glutathione-peroxidase- bound selenium and a slight inverse relation with human serum albumin-bound selenium. Total plasma selenium concentration poorly correlated with most of the selenium species, with the exception of a direct association with selenite and human serum albumin-bound selenium. In gender-specific analyses, some differences emerged about the associations of both dietary and blood total selenium with blood levels of selenium species. Age did not substantially influence these associations. Overall, these results indicate the limited ability of total selenium intake and blood levels in predicting the circulating levels of most selenium compounds. Therefore, epidemiologic studies on the health effects of environmental selenium should focus on the assessment of dietary intake or blood levels of the chemical forms of selenium, and not of total selenium as usually done, in order to avoid substantial misclassification of exposure.